API Reference: Endpoints and Responses

POST/v1/api/transcribe

Queue one video and get a job id back.

Field	Type	What it is
`language_code`	string or null	ISO-639-1/639-3 code of the video's spoken language, e.g. "en"; can improve quality. Omit to auto-detect.
`tag_audio_events`	boolean	Mark non-speech audio with tags like (laughter) or [music]. Off by default: the transcriber picks the tag language itself and does not always match the video's language.
`url`required	string	YouTube video URL
`verbatim`	boolean	Keep filler words and repetitions instead of cleaning them up

Field	Type	What it is
`id`required	string	Job id; poll GET /v1/api/transcription/{id} until `state` is `done`

Answers: 200, 401, 422, 429

GET/v1/api/transcription/{id}

Read a job: its state while it runs, its text when it is done.

Field	Type	What it is
`id`required	string

Answers: 200, 401, 404, 422, 429

POST/v1/api/subtitles

Queue one video and get SRT and WebVTT cues back.

Field	Type	What it is
`format`	string	Which field to read from the result; both are always populated
`language_code`	string or null	ISO-639-1/639-3 code of the video's spoken language, e.g. "en"; can improve quality. Omit to auto-detect.
`lyrics`	boolean	For songs: build the subtitles from the song's published lyrics, timed on the audio. Where several singers sing different lines at once, every line is shown, stacked. Falls back to normal subtitles when no lyrics are found or they do not match what is sung; `lyrics_status` on the result says which happened. Off by default.
`tag_audio_events`	boolean	Mark non-speech audio with tags like (laughter) or [music]. Off by default: the transcriber picks the tag language itself and does not always match the video's language.
`target_languages`	array or null	Also translate into these languages (max 10), e.g. ["Spanish", "French"]
`url`required	string	YouTube video URL
`verbatim`	boolean	Keep filler words and repetitions instead of cleaning them up

Field	Type	What it is
`format`required	string	The requested format, echoed back
`id`required	string	Job id; poll GET /v1/api/transcription/{id} until `state` is `done`

Answers: 200, 401, 422, 429

POST/v1/api/burn-subtitles

Queue one video and get a video with the captions rendered into it.

Field	Type	What it is
`background`	boolean	Draw a semi-transparent black box behind the subtitles for readability. Off = white text with a black outline.
`font_size`	FontSize	Subtitle text size: default ≈7% of video height, large ≈10%, xl ≈12.5%.
`language_code`	string or null	ISO-639-1/639-3 code of the video's spoken language, e.g. "en"; can improve quality. Omit to auto-detect.
`lyrics`	boolean	For songs: build the subtitles from the song's published lyrics, timed on the audio. Where several singers sing different lines at once, every line is shown, stacked. Falls back to normal subtitles when no lyrics are found or they do not match what is sung; `lyrics_status` on the result says which happened. Off by default.
`quality`	Quality	Output resolution: default = 720p, hd = 1080p, uhd = 4K
`tag_audio_events`	boolean	Mark non-speech audio with tags like (laughter) or [music]. Off by default: the transcriber picks the tag language itself and does not always match the video's language.
`target_languages`	array or null	Also produce a burned video per translated language (max 10), e.g. ["Spanish", "French"]
`url`required	string	YouTube video URL
`verbatim`	boolean	Keep filler words and repetitions instead of cleaning them up

Field	Type	What it is
`id`required	string	Job id; poll GET /v1/api/transcription/{id} until `state` is `done`

Answers: 200, 401, 422, 429

The machine-readable version of this page is openapi.json, and it is what this page is generated from.